feat:更新环境依赖，使用uv管理

Author: Liwenru88

src/backend/Dockerfile

@@ -4,8 +4,8 @@ WORKDIR /app
 
 COPY ./ ./
 
-RUN poetry config virtualenvs.create false
-RUN poetry update --without dev
+RUN uv pip install -r <(uv pip compile pyproject.toml --upgrade) --system --no-cache-dir && \
+    uv cache clean
 
 # patch langchain-openai lib. remove this when langchain-openai support reasoning_content
 RUN patch -p1 < /app/bisheng/patches/langchain_openai.patch /usr/local/lib/python3.10/site-packages/langchain_openai/chat_models/base.py

src/backend/base.Dockerfile

@@ -11,9 +11,13 @@ RUN apt-get update && \
     apt-get install -y --no-install-recommends \
     gcc g++ curl build-essential libreoffice \
     wget procps vim fonts-wqy-zenhei \
-    libglib2.0-0 libsm6 libxrender1 libxext6 libgl1 ffmpeg \
+    libglib2.0-0 libsm6 libxrender1 libxext6 libgl1 \
     && rm -rf /var/lib/apt/lists/*
 
+# 安装 FFmpeg
+RUN apt-get install -y ffmpeg && \
+    rm -rf /var/lib/apt/lists/*
+
 
 # 安装 pandoc
 RUN mkdir -p /opt/pandoc && \
@@ -27,16 +31,21 @@ RUN mkdir -p /opt/pandoc && \
 RUN curl -LsSf https://astral.sh/uv/install.sh | sh
 
 # 安装 Poetry
-RUN curl -sSL https://install.python-poetry.org | python3 - --version 1.8.2
+#RUN curl -sSL https://install.python-poetry.org | python3 - --version 1.8.2
 
 # 拷贝项目依赖文件
 COPY ./pyproject.toml ./
 
 # 安装 Python 依赖
 RUN python -m pip install --upgrade pip && \
-    pip install shapely==2.0.1 && \
-    poetry config virtualenvs.create false && \
-    poetry install --no-interaction --no-ansi --without dev
+    uv pip install -r <(uv pip compile pyproject.toml) --system --no-cache-dir && \
+    uv cache clean
+
+
+#RUN python -m pip install --upgrade pip && \
+#    pip install shapely==2.0.1 && \
+#    poetry config virtualenvs.create false && \
+#    poetry install --no-interaction --no-ansi --without dev
 
 # 安装 NLTK 数据
 RUN python -c "import nltk; nltk.download('punkt'); nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger'); nltk.download('averaged_perceptron_tagger_eng')"

src/backend/bisheng/core/ai/rerank/xinference/__init__.py

@@ -0,0 +1,126 @@
+# Copyright 2022-2024 XProbe Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from typing import Dict, Iterable, List, Literal, Optional, Union
+
+from pydantic.version import VERSION as PYDANTIC_VERSION
+
+PYDANTIC_V2 = PYDANTIC_VERSION.startswith("2.")
+
+
+if PYDANTIC_V2:
+    from pydantic.v1 import (  # noqa: F401
+        BaseModel,
+        Field,
+        Protocol,
+        ValidationError,
+        create_model,
+        create_model_from_namedtuple,
+        create_model_from_typeddict,
+        parse_file_as,
+        validate_arguments,
+        validator,
+    )
+    from pydantic.v1.error_wrappers import ErrorWrapper  # noqa: F401
+    from pydantic.v1.parse import load_str_bytes  # noqa: F401
+    from pydantic.v1.types import StrBytes  # noqa: F401
+    from pydantic.v1.utils import ROOT_KEY  # noqa: F401
+else:
+    from pydantic import (  # noqa: F401
+        BaseModel,
+        Field,
+        Protocol,
+        ValidationError,
+        create_model,
+        create_model_from_namedtuple,
+        create_model_from_typeddict,
+        parse_file_as,
+        validate_arguments,
+        validator,
+    )
+    from pydantic.error_wrappers import ErrorWrapper  # noqa: F401
+    from pydantic.parse import load_str_bytes  # noqa: F401
+    from pydantic.types import StrBytes  # noqa: F401
+    from pydantic.utils import ROOT_KEY  # noqa: F401
+
+from openai.types.chat.chat_completion_named_tool_choice_param import (
+    ChatCompletionNamedToolChoiceParam,
+)
+from openai.types.chat.chat_completion_stream_options_param import (
+    ChatCompletionStreamOptionsParam,
+)
+from openai.types.chat.chat_completion_tool_param import ChatCompletionToolParam
+from openai.types.shared_params.response_format_json_object import (
+    ResponseFormatJSONObject,
+)
+from openai.types.shared_params.response_format_text import ResponseFormatText
+
+OpenAIChatCompletionStreamOptionsParam = create_model_from_typeddict(
+    ChatCompletionStreamOptionsParam
+)
+OpenAIChatCompletionToolParam = create_model_from_typeddict(ChatCompletionToolParam)
+OpenAIChatCompletionNamedToolChoiceParam = create_model_from_typeddict(
+    ChatCompletionNamedToolChoiceParam
+)
+from openai._types import Body
+
+
+class JSONSchema(BaseModel):
+    name: str
+    description: Optional[str] = None
+    schema_: Optional[Dict[str, object]] = Field(alias="schema", default=None)
+    strict: Optional[bool] = None
+
+
+class ResponseFormatJSONSchema(BaseModel):
+    json_schema: JSONSchema
+    type: Literal["json_schema"]
+
+
+ResponseFormat = Union[
+    ResponseFormatText, ResponseFormatJSONObject, ResponseFormatJSONSchema
+]
+
+
+class CreateChatCompletionOpenAI(BaseModel):
+    """
+    Comes from source code: https://github.com/openai/openai-python/blob/main/src/openai/types/chat/completion_create_params.py
+    """
+
+    messages: List[Dict]
+    model: str
+    frequency_penalty: Optional[float]
+    logit_bias: Optional[Dict[str, int]]
+    logprobs: Optional[bool]
+    max_completion_tokens: Optional[int]
+    max_tokens: Optional[int]
+    n: Optional[int]
+    parallel_tool_calls: Optional[bool]
+    presence_penalty: Optional[float]
+    response_format: Optional[ResponseFormat]
+    seed: Optional[int]
+    service_tier: Optional[Literal["auto", "default"]]
+    stop: Union[Optional[str], List[str]]
+    stream_options: Optional[OpenAIChatCompletionStreamOptionsParam]  # type: ignore
+    temperature: Optional[float]
+    tool_choice: Optional[  # type: ignore
+        Union[
+            Literal["none", "auto", "required"],
+            OpenAIChatCompletionNamedToolChoiceParam,
+        ]
+    ]
+    tools: Optional[Iterable[OpenAIChatCompletionToolParam]]  # type: ignore
+    top_logprobs: Optional[int]
+    top_p: Optional[float]
+    extra_body: Optional[Body]
+    user: Optional[str]

src/backend/bisheng/core/ai/rerank/xinference/_compat.py

@@ -0,0 +1,20 @@
+# Copyright 2022-2023 XProbe Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from .restful.async_restful_client import AsyncClient
+from .restful.restful_client import Client
+
+# For compatibility
+RESTfulClient = Client
+AsyncRESTfulClient = AsyncClient

src/backend/bisheng/core/ai/rerank/xinference/client/__init__.py

@@ -0,0 +1,99 @@
+# Copyright 2022-2023 XProbe Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import json
+from typing import Any, AsyncIterator, Iterator, Union
+
+
+def convert_float_to_int_or_str(model_size: float) -> Union[int, str]:
+    """convert float to int or string
+
+    if float can be presented as int, convert it to int, otherwise convert it to string
+    """
+    if int(model_size) == model_size:
+        return int(model_size)
+    else:
+        return str(model_size)
+
+
+def streaming_response_iterator(
+    response_lines: Iterator[bytes],
+) -> Iterator[Any]:
+    """
+    Create an Iterator to handle the streaming type of generation.
+
+    Note
+    ----------
+    This method is for compatible with openai. Please refer to:
+    https://github.com/openai/openai-python/blob/v0.28.1/openai/api_requestor.py#L99
+
+    Parameters
+    ----------
+    response_lines: Iterator[bytes]
+        Generated lines by the Model Generator.
+
+    Returns
+    -------
+    Iterator["CompletionChunk"]
+        Iterator of CompletionChunks generated by models.
+
+    """
+
+    for line in response_lines:
+        line = line.strip()
+        if line.startswith(b"data:"):
+            json_str = line[len(b"data:") :].strip()
+            if json_str == b"[DONE]":
+                continue
+            data = json.loads(json_str.decode("utf-8"))
+            error = data.get("error", None)
+            if error is not None:
+                raise Exception(str(error))
+            yield data
+
+
+async def async_streaming_response_iterator(
+    response_lines: AsyncIterator[bytes],
+) -> AsyncIterator[Any]:
+    """
+    Create an AsyncIterator to handle the streaming type of generation.
+
+    Note
+    ----------
+    This method is for compatible with openai. Please refer to:
+    https://github.com/openai/openai-python/blob/v0.28.1/openai/api_requestor.py#L99
+
+    Parameters
+    ----------
+    response_lines: AsyncIterator[bytes]
+        Generated lines by the Model Generator.
+
+    Returns
+    -------
+    AsyncIterator["CompletionChunk"]
+        AsyncIterator of CompletionChunks generated by models.
+
+    """
+
+    async for line in response_lines:
+        line = line.strip()
+        if line.startswith(b"data:"):
+            json_str = line[len(b"data:") :].strip()
+            if json_str == b"[DONE]":
+                continue
+            data = json.loads(json_str.decode("utf-8"))
+            error = data.get("error", None)
+            if error is not None:
+                raise Exception(str(error))
+            yield data

src/backend/bisheng/core/ai/rerank/xinference/client/common.py

@@ -0,0 +1,36 @@
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulAudioModelHandle as AsyncAudioModelHandle,
+)
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulChatModelHandle as AsyncChatModelHandle,
+)
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulEmbeddingModelHandle as AsyncEmbeddingModelHandle,
+)
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulGenerateModelHandle as AsyncGenerateModelHandle,
+)
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulImageModelHandle as AsyncImageModelHandle,
+)
+from .restful.async_restful_client import (  # noqa: F401
+    AsyncRESTfulVideoModelHandle as AsyncVideoModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulAudioModelHandle as AudioModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulChatModelHandle as ChatModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulEmbeddingModelHandle as EmbeddingModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulGenerateModelHandle as GenerateModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulImageModelHandle as ImageModelHandle,
+)
+from .restful.restful_client import (  # noqa: F401
+    RESTfulVideoModelHandle as VideoModelHandle,
+)

src/backend/bisheng/core/ai/rerank/xinference/client/handlers.py

@@ -0,0 +1,13 @@
+# Copyright 2022-2023 XProbe Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.